REVIEW 6 cited by
Towards Stable and Efficient Training of Verifiably Robust Neural Networks
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
abstract
Training neural networks with verifiable robustness guarantees is challenging. Several existing approaches utilize linear relaxation based neural network output bounds under perturbation, but they can slow down training by a factor of hundreds depending on the underlying network architectures. Meanwhile, interval bound propagation (IBP) based training is efficient and significantly outperforms linear relaxation based methods on many tasks, yet it may suffer from stability issues since the bounds are much looser especially at the beginning of training. In this paper, we propose a new certified adversarial training method, CROWN-IBP, by combining the fast IBP bounds in a forward bounding pass and a tight linear relaxation based bound, CROWN, in a backward bounding pass. CROWN-IBP is computationally efficient and consistently outperforms IBP baselines on training verifiably robust neural networks. We conduct large scale experiments on MNIST and CIFAR datasets, and outperform all previous linear relaxation and bound propagation based certified defenses in $\ell_\infty$ robustness. Notably, we achieve 7.02% verified test error on MNIST at $\epsilon=0.3$, and 66.94% on CIFAR-10 with $\epsilon=8/255$. Code is available at https://github.com/deepmind/interval-bound-propagation (TensorFlow) and https://github.com/huanzhang12/CROWN-IBP (PyTorch).
Forward citations
Cited by 6 Pith papers
-
Certified Training for Convolutional Perturbations
A certified-training method using parameterized blur kernels and symbolic bound propagation gives provable robustness to motion blur and related convolutional perturbations, reaching over 80% verified accuracy on CIFAR10.
-
SAIL: Sound Abstract Interpreters with LLMs
SAIL synthesizes globally sound abstract transformers for neural-network operators by combining LLM generation with syntactic validation, SMT-based soundness checking, and cost-guided iterative refinement.
-
Robustifying Diffusion-Denoised Smoothing Against Covariate Shift
Adversarially perturbing the noise term of a diffusion denoiser during training improves the certified l2 robustness of denoised randomized smoothing on MNIST, CIFAR-10, and ImageNet, with the largest gains at large p...
-
Compression Aware Certified Training
CACTUS trains a single network on pruned and weight-perturbed copies of itself, beating prior certified-training baselines on compressed MNIST and CIFAR-10 models.
-
TriGuard: Testing Model Safety with Attribution Entropy, Verification, and Drift
TriGuard reports that attribution drift and entropy provide safety insights orthogonal to adversarial accuracy, and that entropy-regularized training reduces this drift.
-
Learning to Optimize by Differentiable Programming
A tutorial survey of differentiable-programming-based first-order optimization, with dual-based PyTorch case studies and no new results.
Discussion (0). Continue with ORCID to comment.